Recap last lecture

  • from unique words to comparable embeddings 🧮
    • relational meaning
    • more granular representations are more effective
  • modern, data-driven NLP is both powerful and biased 🚨
    • there is nothing like raw data or objective models

Outline

  • get some organizational stuff done
  • let’s do serious NLP! ✨
  • code interactively
    • interrupt, ask, and complement

REMINDER: Course Evaluation 📣


Thanks for any constructive feedback,
be it sweet or sour! 🙏

Your mini-projects 📌

You are ready to work (self-paced)

  • your project idea is recorded here
  • see requirements here
    • group size max. 3 people
  • find inspiring projects of the past here


Happy to help! Reach out if you are stuck! 🤯

Let’s apply this in practice ✨